Computers are still no substitute for wind tunnels, which have turned play-maker in shaping the future home of London's Arsenal Football Club. Mark Hansford reports.
Uk premiership champion, cup winner, and now a new stadium on the way. Quite a treble for London's Arsenal Football Club. But while imported talent in the form of overseas players can take the credit for the club's clean sweep of UK soccer trophies, it is a truly British affair playing a key role off the field.
BMT Fluid Mechanics is a seasoned player, dating back to the days of nationalised research, and the UK's National Physics Laboratory. Some 20 years post-privatisation, it is still an anchor at the back, passing invaluable data on wind-loading and wind-effects forward to the designers and architects of the spectacular new 60,000 seat stadium at Ashburton Grove in north London.
BMT's involvement in the Arsenal project has been typical for this kind of structure, explains manager of fluid and structural mechanics Volker Buttgereit. It began with the environmental impact assessment at planning stage and carried through to detailed design.
'Pedestrian level wind dispersal is now a major part of any environmental statement, ' says Buttgereit. Tall buildings generate strong vortex flow, and the effects can start up to 100m away. They can also be hard to dissipate. 'Vortices from 747 jets are smaller than those from tall buildings, yet can remain on the runway for 20 minutes after take-off, ' Buttgereit explains.
Fortunately for Arsenal, stadia impact on wind far less than tall buildings, which are much bluffer obstructions with significantly stronger down-draughts.
BMT was far more interested in questions that affect detailed design, Buttgereit explains, namely, the likely wind loading on the roof structure and cladding, and environmental flow such as wind flows at pitch level and wind-driven rain.
'The unusual shape of the structure means there is nothing in BS 6399. The Eurocode, which is just a compilation of random pieces of wind research, is still under review. And the code of practice doesn't allow for the effects of adjacent buildings, ' says Buttgereit. 'So where would you be without a wind tunnel?'
BMT's analyses are not, however, limited to wind tunnels.
Computational fluid dynamic (CFD) modelling also plays a major role, particularly with environmental flow modelling.
'Scaling effects mean that some aspects of wind flow are not well represented in the wind tunnel, ' says Buttgereit. 'CFD is particularly useful for thermal effects, looking at spectator comfort.'
CFD is also used to examine what is now a common problem with modern enclosed stadia - grass growth. 'The roof design is important for internal flow. So we plot velocity at pitch level from different wind angles and use it to advise the design team, ' explains Buttgereit. 'But ultimately you can't make air flow against a thermal gradient and the best way to solve the problem is to introduce cold air at ground level.'
In the wind tunnel, upstream roughness boards model turbulence and 500 pressure taps attached to the stadium model measure the loading. Transducers translate the pneumatic signals into electronic ones for recording by computer. In any one day three wind directions can be simulated at 36 different wind speeds, all of which are fed into the computer before results can be declared with confidence. For Arsenal, Buttgereit expects to reduce code factors of safety from 2.0 to between 1.2 and 1.4, a significant saving in overdesign.
'For cladding on medium to low rise buildings we can offer cost savings of 30%, ' Buttgereit claims.
Such benefits are now resulting in a miniboom at BMT, which it also puts down to 'industry efficiency drives and our own slickness. In the past the Arsenal Stadium would have taken months, now we can do it in a couple of weeks.
'We are almost under pressure to deliver a conveyor belt approach, ' he adds. 'The trick is knowing when to hit the emergency stop and say 'there's something wrong here'.'
INFOPLUS www. bmtfm. com
The case for wind tunnels
Wind tunnel testing at the National Physics Laboratory in Teddington, west London, dates back to pre-war times. In the immediate post-war period, 60 wind tunnels provided invaluable data for the engineering boom. While demand may have dipped, wind tunnels at the now-privatised BMT Fluid Mechanics are still providing data that can be found nowhere else.
'The fundamental problem is solving the turbulence, which has been the case for 40 years since the Navier-Stokes equations were established, ' explains Buttgereit.
These non-linear partial differential equations, universally applicable in all flow types, have one major drawback in turbulent flow - such as that experienced around stadia - in that simple mathematical solutions are impossible.
'Turbulence modelling defines approximations to the real terms in N-S, ' Buttgereit explains. 'But it rarely works for the full range, from localised effects to the whole structure. Experience is the most useful asset.'
'With a wind tunnel, what you see is what you get. With CFD (computational fluid dynamics) it is much more difficult to see the difference between model errors and physical actions, ' says Buttgereit.
Used correctly, CFD can No nonsense: 'With wind tunnel, what you see is what you get.'
have important role to play. But according to Buttgereit: 'The difficulty with flow effects is that you can always think of a post-rational explanation for what may, in fact, be a wrong result, ' he goes on. 'Unless you know the correct parameters to use the answer given will be wrong.'
Buttgereit cites a recent test where three experienced users were challenged to use a commercial CFD tool on a simple cube. The mean force coefficients calculated varied by more than 200%. 'CFD needs to be viewed with caution, ' says Buttgereit.
'With respect to wind everybody thinks that they are an expert. People rarely come to us without preconceptions, ' bemoans Buttgereit.
'Yet I've spent my whole life doing it, and can't understand it without a wind tunnel model.
That's why we're doing the work we do.'